I love fiction. UFOs, unicorns, faith healing, the Mobile Web, the Mobile Context, psychics. Speaking of which…
I’m sure many of you have heard of cold reading. Cold reading is one of many toolkits used by psychics to convince others (and often themselves) that they know more about a person than they actually do. One of the more well-known and obvious cold reading techniques is the use of simple, broad-stroke statistics. There are lots of people in the world named John. Chances are, if I looked you in the eye, contorted my face, feigned mind-agony and said, “I see a J. Someone in your life… James? John? Someone is having some serious financial trouble,” I’d have a good chance of being technically correct. And that’s not even a good read. But the thing that makes ruses like this work is that you, the user, are filling in the blanks for me. You search your brain and you find John. Or James. Oh you mean Jason, my cousin’s sister’s boyfriend, who does have financial trouble.” Two weeks later, you’ll swear I knew about Jason’s stolen credit card.
In the Mobile Context—you know, that magical fairy-tale place we go whenever we use the Web on our tablets or phones—users apparently don’t need to fill in the blanks. The People Who Make Websites make these decisions. In the Mobile Context, you see, people change. The slide lock on their phone transforms them into beings with different needs. And I use an Android phone, which means my needs are different than an iOS user, because the user experience is obviously not as important to me. Cough.
Cold reading happens in the Mobile Context. Some might find the above sarcasm a rehash of Mark Kirby’s wonderful mind reading assertion. But it’s not, because if I could really mind-read, I would really know what you’re thinking. But I don’t. So I’m not mind-reading, I’m cold reading. And you’re not able to fill in the blanks for me, because a web page is mostly a one-way thing. I can’t quickly change my story by watching your face for signs that I’m on the right track. So when I decide to make decisions for you, I’m potentially sabotaging the experience for the both of us.
So what do we Web-makers do? We infer things based on the device you’re using. And statistics. Analytics. Which means we’ll probably be right much of the time, or some of the time. But there is someone who is right all of the time when it comes to knowing which task a user wants to do in a given situation, even in pretend-universes like Mobile Context: the user, right? If I walk into a coffee shop where they also sell sandwiches, you might be wrong to assume I want a coffee. But if you ask me what I want, and I say, “I’d like a sandwich,” there’s a pretty decent chance that I want a sandwich, yes? And you didn’t get distracted by the fact that 87% of your clients order coffee. And the most beautiful part is this: you didn’t have to put up new menus and change the furniture because I’m a sandwich person. You just need to offer me the choice of a sandwich or a coffee and guess what? I’ll make the choice and order my sandwich.
At Mobilism this year I met Christiaan Lustig, a nice guy and web professional who’s doing a lot of research into responsive design related subjects. He recently wrote an article about top task analysis. The article was actually about many things and it seems so fair to each point of view that I may be missing a single defined argument, but prioritizing top tasks based on user’s needs seems to be the gist of the article. I can agree with that. However, the article also used some examples that really show some current industry thinking I strongly disagree with. Here’s an excerpt from an example about a bank site (go take a look) that just kills me:
“More than 50 per cent of their website visitors are on a mobile device. And 98 per cent of those users go directly to the online banking login.”
That means 49 per cent of all visitors go directly to the online banking login.[…]
First, take a minute to recover from the shock that a majority of banking customers actually plan to do some banking on a bank website. Then notice that we’re provided with only the mobile side of the statistics. 49 percent of all visitors is just plain wrong, unless none of the desktop users go directly to the banking login. We can’t learn from conclusions unless we know what led to them. What do the desktop users do on the site?
To be clear, I’m not criticizing the author of the article at all. I am criticizing the thinking and assumptions behind what many organizations are doing in and around content and responsive web design.
Looking at the screenshots next to the article’s bank example, you can’t help but agree with the author that the mobile rendering of the site seems to be more task oriented. But is that because desktop users really want stupid fluff over tasks? Or is the mobile version of the page the only one for which the Web People seriously considered the needs of the user? In other words, does content really need to change for a mobile context, or do we simply need better content all across the board? I bet the desktop site would be better if they just chucked it and used the “mobile” site for everything. Is the mobile view better because it’s better? Or is it simply better in contrast to a sucky desktop view with irrelevant content?
This example whittles existing content down to a small series of possible tasks on the mobile view. If these are all the tasks a user can do on the site, great. If not, why not? Is it cold in here? Is it cold reading in here?
Let me make a long story short: just make quality, relevant content with appropriate tasks, and offer all of these to all users, unless said content or functionality is dependent upon device capabilities (such as a camera). Then make it easy for the user to decide what it is they want to do. It’s like… web development all over again, isn’t it?
I really don’t see any reasons for the content of a page to change completely under any device-related circumstances. In fact, for purposes of clarity and findability, why not go out on a limb and actually use URLs to refer to specific content or functionality, instead of trying to morph homepage content into something different on a phone and hoping you get the assumptions right? example.com/login, example.com/plan, example.com/status. You’re allowed to have a page which simply contains the tasks and content it contains. Really, it’s okay. It’s responsive design; we’re not making shapeshifters.
Alter the design as you will. Visually, content order is most likely linear on phones, so you’ll account for that. Text size, link and button size, image (file) size… there are plenty of things to consider. Of course you might not be able to have that Really Important Canvas Animation. Of course you’ll want to be able to use some device capabilities like your phone’s camera if that’s available. That’s logical and fine. But if something is essential content, functionality, or common sense on all platforms and you’re only providing it on some, you’re doing it wrong, unless your site or app is unavailable on other platforms.
Users are ultimately the experts in deciding what they want to do on your site. Make it possible, make it easy, then get out of the way and let them fill in the blanks. They’ll attribute the good experience to you.
I think you are mistaking flawed arguments for a flawed framework. There are some very stable truths about the way users interact with different devices. Such as their affinity to luxury brands, and the types of tasks that drive their visit.
I have always seen, thus far, a much clearer “need” in the web-traffic for mobile users over the desktop users of the same site. Even the relatively new class of tablet users show a leaning toward high-interactivity instead of task completion.
The web is not monolithic; there are variety of loosely connected webs that make up sections of the Internet as a whole. I think the best analogy is the Web is to Internet as Galaxy is to Universe.
“I don’t always blog, but when I do, it’s fucking brilliant.” -Stephen Hay, the most interesting man in the (mobile) world
Great article that articulates something that has been hard to define and argue about for the last couple years.
I totally agree with you regarding making content easily accessible regardless of context, and also the importance of URLs (something I hope to write about soon). Spot on. Users aren’t so stupid that they don’t know to click a link in order to get to where they need to go. Those types of contextual “what ifs” can be solved simply by making links to features accessible.
However, I pose this scenario to you. Let’s say you’re eBay, and you see desktop users on average buying $100 worth of stuff typically in apparel, furniture and gardening. Mobile phone users, on the other hand, spend about $60 typically in electronics, video games and music.
Now it’s time to design the homepage. Of course we make sure all categories are quickly accessible and no content is being arbitrarily cut on mobile (Who would ever do that?! Apparently, everyone everywhere). But now it comes time to choose which products to feature on the homepage. Do you serve the same products to all users, knowing full well that mobile users statistically prefer other categories than apparel, furniture and gardening? I would say it would be in eBay’s best interest to conditionally serve up different featured content to mobile users than to desktop users. Higher visibility of key mobile products and categories potentially means more mobile conversions means more money.
It’s still a guessing game and you can’t assume that *all* mobile users are only interested in electronics, video games and music. But that’s OK provided you still give access to the other categories. But I do think it makes logical sense to serve different homepage content for different contexts in that instance. I see it as more “enhancement and optimization” rather than mind-reading.
I’d love to hear your thoughts. Thanks again for a great article!
Good points all, and I’m glad the word’s getting ’round that the interface may change, but the actual content and tasks available to mobiles should not. Usually that’s taken to mean that we should bring the desktop experience to the mobile, but it’s an astute observation that mobile may have more to teach us about desktop at this point.
Cutting fluff from designs is a goal I need to set for myself. If anybody pioneered that back in the day, I suppose it would have been Google, but look at where we are now. The simple, pure search engine page has given way to Google+ and dozens of other services, adorable hedgehog wallpaper, rich snippets, ads, a sidebar with fourteen different search types, &c., &c., &c. Besides which, I wonder whether SEO hasn’t driven much of the cruft on our own websites—tags, categories, authors, related posts, highest-rated posts, most-commented posts, posts my pet rabbit enjoyed…all washed down with a slurry of keywords and stilted content.
Hopefully we can start walking this back and begin to design simpler, more useful sites for all platforms. Less, but better.
@carlos: What point are you trying to make, exactly?
@brad: What you’re talking about is something totally different than what I’m talking about. You’re referring to one of the things that Christiaan referred to in his article: prioritization. You’re *emphasizing* (read: not changing or replacing it) something based on platform-specific data. I see no problem with that, as you’re not changing basic content, tasks or functionality. Featured content, such as in your example, *usually* rotates or gets regularly changed or updated, so I think most users would not be bothered by that at all if the other choices were clear and easy to get to. You’re turning the A/C on in a car automatically because it’s hot outside, not replacing the steering wheel with a big A/C button.
Stephen, thanks for you article. Brilliant stuff, like Brad said!
I’ll keep thinking things through regarding top tasks and responsive web design, and what that could mean to prioritization and — if and when applicable ;-) — cutting content and features from mobile/tablet/desktop views.
Thanks again, and I hope to see you soon!
The most notable place where I suffer from what you pointed out is the
google mobile site. I have a basic phone with a basic browser, and for that matter, I can’t use those cool search filters. There is, indeed, an option to switch to the desktop ( option to switch to the full desktop site on responsive designs: must !important;) , but I have a basic phone with a basic browser, again, so it’s weird. There is a link to the search settings page, but that too is mobile over-optimized and very limited. A link to the advanced search page should have helped me a lot.
Provide the same content, on all platforms(conditions apply). while prioritizing it according to some common usage patterns and utilizing some certain features of a platform(as you mentioned, using the camera on mobile devices and utilizing the large screen size, advanced browser features cpu power .etc on Desktop) will be the best thing that we can do.
Your thesis: (context = cold reading) is funny, but wrong. Your implied sub-point (there is only one web) I disagree with.
I am onboard with Brad’s assertion that site specific segmentation is meaningful. In many situations tailored content is more effective than shoehorning visitors onto a single page that is meant to respond to all devices. Seperation of desktop and mobile users into entirely different versions of the site in many ways supports a web makers ability to focus on being useful.
@Carlos: “Cold reading = context” might have been wrong had I said it, but I did not. That said, your disagreement with my point of view does not make it wrong. `4 + 7 = 3` is wrong.
I also don’t agree that Brad’s assertion is what you claim it to be. Apparently we’re both interpreting his comment differently.
If your point is that you disagree with my point of view, that’s fine. I have great respect for that. However, I’ll only consider myself wrong when I stop seeing frustrated users “going to the Desktop Version” in order to get to a piece of expected base content or functionality.
I accept that I should have said disagree, not wrong. I will add that some users prefer the Desktop not because it has a different functionality, but because it has a familiar structure that leaves them with nothing to think about. Particularly when someone know a site intimately it is perceived faster to use the familiar interface, in spite of the the often more functional presentation of mobile specific content.
I am interested to know what your reading of Brad’s comment?
@Carlos: I agree with you that many users prefer the familiar. Absolutely. An example of what I’m referring to is Mahroof Ali’s comment above, where he expects the same functionality to at least be available in the mobile version (although it may be presented in a way that is optimized for the device).
I interpreted Brad’s comment as using hard data about mobile use to justify giving certain content a higher priority (and also different presentation) on a mobile device, all the while still making it easy to get to content users might expect due to their use of the site on the desktop. He referred to a portion of the content, not necessarily the whole of it (which might justify use of different URL).
That was my take on it. I think his approach removes the *need* for users to ever visit the “desktop version”, although I tend to agree with you that some always will want to.
@Stephen you’re right about my interpretation. I’d say a good rule of thumb is that you should serve the same content coupled with an optimized presentation unless you have hard data and a really unique use case to justify otherwise.
@Carlos People do prefer familiar, but I can make a pretty good claim that most current mobile web users who prefer a desktop experience is because they are confronted with terrible mobile web experiences that don’t give them what they’re looking for. As mobile optimized experiences become more robust and fully-featured, we’ll start seeing a dramatic decrease of clicks to “view full site”.
@ Brad:
“I’d say a good rule of thumb is that you should serve the same content coupled with an optimized presentation unless you have hard data and a really unique use case to justify otherwise.”
Fully agreed. This was the point I was trying to make in my article. I tried to argue that Gerry McGovern’s top task analysis method gives you that hard data.
Finally got around to reading this. Nice article.
With regards to your response to Brads original comment…
“You’re referring to one of the things that Christiaan referred to in his article: prioritization. You’re *emphasizing* (read: not changing or replacing it) something based on platform-specific data. I see no problem with that, as you’re not changing basic content, tasks or functionality. Featured content, such as in your example, *usually* rotates or gets regularly changed or updated, so I think most users would not be bothered by that at all if the other choices were clear and easy to get to.”
… I had actually read it in another way. For sales driven sites, and lets take the ebay example given, I think you would encourage the replacement or change of the featured items on the home page with access to the other categories if the user was more interested rather then putting them on a carousel.
I would think the same thing would go for accessing an airline website for flights. The homepage would load the deals flying from the country you were situated in only, and then provide the usual links to alternative deals from other departure locations. You’re not taking anything away from the user, your just making the content more relevant to what you know about the user.
These are specifics though ( probably nit picking, sorry) and what i think should be considered exceptions to the rules that you have described in your article because for the majority of websites out there I think you’ve hit the nail on the head.
@Justin: I agree with you for the most part (please note that Brad confirmed my interpretation of his comment), and I think people should be careful with that last thing you said,
“You’re not taking anything away from the user, your just making the content more relevant to what you know about the user.”
That seems to make sense. However the fact that you know *something* about the user (their location, in this case) does not mean that your content choices based on that fact are more relevant to them. That’s trying to make a connection between what *you* know about the user and what the user is thinking. Sometimes it’ll work, but sometimes it won’t. So in that case, I applaud your choice to not *remove* any content, just in case your location-based assumption is incorrect for some users. Thanks for your thoughts!
I’m a sucker for a good context conversation…
The only additional consideration I see in @Brad’s approach (eBay featured content example) is that this kind of decision must be continually monitored. To be more specific, because a site’s analytics state something like, “visits from small screens lead to a majority of purchases in electronics, etc” is great, but it’s not the small screens that are making the purchases, it’s the owners of those small screens. So, perhaps at this point in time, more younger users may be browsing eBay with a smaller screen while more older users are browsing with larger screens. The real data is that a younger audience spends their doe on electronics and NOT furniture. These analytics can be a little misleading if we don’t keep the users in mind.
Obviously, all of this is hypothetical, but if there was an influx of older users getting online with small-screened devices, that stat (which is pulled from the history—as in the past) may be quickly out of date.
I really do agree that users generally want/expect the same content. But I also think we have to find ways to gather more clues about their context. Content prioritization (like the two links at the top of the http://mobilism.nl/ site itself at smaller resolutions) could be one way to take action with those clues. I just can’t believe that we’ve explored this topic enough. I’ve seen a lot of talk lately about allowing our users to define their own context, which you hint at in the article above. I think there’s much progress to be made with this…
Great stuff Stephen, keep it up!
The funny thing about that “98% of mobile users go to online banking” bit is that when I worked at a credit union, something like 80% or more of our desktop visitors (and this was 5 or 6 years ago) went directly to online banking. So when we redesigned, we made the login really easy to find, while still giving access to the rest of the site, and having some eye-catching marketing bits. (We also did some usability testing that gave us similar info.)